山脊
涡轮机
跟踪(教育)
海洋工程
风速
计算机科学
控制理论(社会学)
环境科学
汽车工程
工程类
地质学
机械工程
人工智能
海洋学
控制(管理)
心理学
古生物学
教育学
作者
Changliang Liu,Shaokang Liu,Yingjie Wu,Weiliang Liu,Shuai Liu
摘要
The instantaneous shaft speed estimation method of wind turbine gearbox based on time-frequency ridge estimation is the most widely used method, but it requires parameter adjustment and expert knowledge to ensure the accuracy of ridge estimation, which significantly reduces the effectiveness of such methods in practice. In this paper, an end-to-end automated ridge estimating (ARE) method is proposed. ARE uses the inherent harmonic relationship in the vibration signal of wind turbine (WT) gearbox to construct a selection index for automatically selecting the target ridge with clear physical meaning. Then, a sub-region division strategy is proposed to avoid the interference of other components and improve computational efficiency. Finally, a parameter-free objective function considering both the amplitude and smoothness of the ridge is proposed to achieve target ridge estimation. The vibration signals of different WT gearboxes collected on-site were applied to verify the performance of ARE. The result shows that ARE realizes the automatic estimation of a ridge. It outperforms other methods in terms of accuracy and computational efficiency.
科研通智能强力驱动
Strongly Powered by AbleSci AI